MIMO OFDM Channel Estimation with Optimum Pilot Patterns for Cognitive Radio in Overlay Spectrum Sharing System
H. Abdulrahman · Research Repository (Delft University of Technology) · 2009
The demand for high data rates in wireless communication systems is increasing as well the demand for new (communication) services. Expanding of existing services or introducing new services needs a portion in the radio spectrum. Mean while the radio spectrum is a scarce resource and is regulated by governmental regulations and policies. Measurements done by the FCC show that the licensed spectrum is not always fully used and hence is not efficiently utilized. This need for spectrum resources and the unefficient use of it has lead to the birth of Cognitive Radio (CR). CR enables the efficient use of the radio spectrum and makes it possible for new services to operate adjacent to existing Licensed Users (LU) without suppressing their performance. There are several methodologies to realize CR, one of those methodologies is Orthogonal Frequency Division Multiplexing OFDM. OFDM is one of those methods which is very suited to be used as a modulation and transmission technique in a CR system. OFDM is suited due to its multicarrier property and its robustness in the faded mobile wireless environments. In almost all wireless applications the arrival of the transmitted message in a good and acceptable quality is an agreement condition. Hence the Bit Error Rate (BER) is important and must be boosted. Multiple Input Multiple Output (MIMO) has been known as a technique to enhance the transmission quality and capacity. Learning and understanding capabilities in CR are crucial to achieve the goals of efficient spectrum utilization in terms of spectrum resources and energy resources. The wireless channel is central within this context, thus estimating the channel is the key to make CR operational, taking in consideration that the transmission-reception technology is available. In this thesis, we design a MIMO system using OFDM modulation technology to transmit and receive two signals over the mobile wireless channel. We use MIMO concept to enhance system capacity and robustness of the wireless transmission. Due to the influence of the channel on the transmitted messages and its importance in wireless transmission to reconstruct the transmitted signals, the channel needs to be known as much as possible. We gather knowledge about this wireless channel and try to estimate it using training symbols called pilot symbols. We distribute the pilot symbols in a way were they can estimate the channel. The way those pilots are organized and implemented in the data stream is called the pilot pattern. We design three different pilot patterns suitable for implementation in MIMO. Within the pilot patterns, we implement the novel Virtual Pilot concept, to simplify the filtering process and to save energy, because the virtual pilots don’t consume power. We test all three patterns. Based on their performance in estimating the channel, we chose the best performing one and implement it in aMIMO OFDMbased CR system. The performance of each pattern is measured by means of BER and Mean Square Error (MSE). We assume that information about the LU is available from spectrum sensing for the transmitter. Due to the properties of the OFDM frame (2D), the estimation is done by using Wiener filtering first in frequency direction followed by filtering in time direction. To reduce computation complexity of the filtering process, we chose the cascaded Wiener filtering as 2x1D filtering instead of 2D filtering, because the first have a comparable performance compared with the last and have less computational complexity. Due to the MIMO structure, the received signal is the contributions from all transmit antennas. Those different contributions from the transmit antennas at the receiver need to be separated. To separate the different contributions form each other, we use the layering concept. Vertical Bell Laboratories Layered Space-Time (V-BLAST) is a detection algorithm to be used in each receiver to detect and separate the transmitted signals in the received stream from each other.